): Cortical function relies on complex neuronal circuits composed of multiple cell types and millions of connecting synapses. During late development, immature cortical networks are progressively optimized via a process known as synaptic pruning. During this period, neurons are particularly sensitive to external stimuli, and activity- dependent signals guide circuit refinement through selective stabilization or elimination of specific synaptic connections. Elucidating the molecular mechanisms underlying activity-dependent synapse selection is key to understanding this fundamental aspect of brain development and plasticity. Activity-regulated genes are prime molecular applicants for mediating the effects of neuronal activity on synapse formation and elimination. One example is applicant plasticity gene 15 (CPG15), which encodes a small glycosylphosphatidylinositol (GPI)- linked protein attached to the cell surface. CPG15 has been previously implicated in axonal and dendritic growth and the maturation of excitatory synapses. New preliminary results reveal that CPG15 knockdown reduces recruitment of the postsynaptic density protein 95 (PSD95) to newly formed dendritic spines of pyramidal neurons, thus reducing spine stabilization. This finding is mechanistically puzzling given that CPG15 is extracellular while PSD95 is intracellular, and neither protein possesses a transmembrane domain. Recently, investigation of the AMPA-type glutamate receptor proteome identified CPG15 as part of the protein complex that co-precipitates with AMPA receptor subunits. To test the role of CPG15 in spine stabilization, our goal is to investigate the effect of CPG15 direct interaction with AMPA receptors to mediate the recruitment of PSD95 to newly formed excitatory synapses. With the power of combined in vitro and in vivo approaches, our data can reveal, for the first time and in unprecedented detail, the timing and sequence of AMPAR and PSD95 recruitment and shed light onto molecular signals that control the formation and continuous adaptation of neuronal networks. The in vitro and in vivo analyses of the CPG15 knockout mouse could easily be translated to a KO model for any neurodevelopmental or neurodegenerative disorder and provide crucial training in in vivo imaging, molecular tool development, and sophisticated data analysis, all applicable to my future work as an independent researcher. Training with Dr. Nedivi, my sponsor and an internationally recognized expert in the field of synaptic plasticity and in vivo 2-photon imaging, is an ideal fit for the project proposed and for my career goals as outlined in the training plan. Being an active member of the vibrant neuroscience community at MIT and Boston, will help me establish scientific relationships with leaders in the field and postdoctoral fellows who will become my colleagues in the future. The Nedivi lab, the Picower Institute for Learning and Memory, and MIT foster innovative work at the interface of neuroscience and engineering, an ideal fit for my academic growth. This outstanding postdoctoral training environment will facilitate my success in launching a career as an independent investigator.